Wednesday, February 22, 2012

Under the Hood: HTML5 or Native? A Guide

This post was also published on CNET and VentureBeat.

Taking your site mobile is a technology minefield. Here's how we're doing it at CBS Interactive.

The mobile technology landscape is incredibly confusing. There are numerous choices, ranging from new HTML5 technologies, native app development methods, and all sorts of content management systems.
At CBS Interactive, we have numerous mobile solutions, including native apps for, CNET, and "60 Minutes," along with mobile-optimized Web sites for GameFaqs and global properties like ZDnet.

At first blush, it seems problematic that various properties have picked completely different architectures for mobile delivery. A technologist's initial inclination is to have everyone run a consistent architecture across all of our properties. Yet it actually makes sense to run a variety of architectures to support mobile delivery.

The biggest issue to address is the ongoing tension between HTML5 and native. Most of the debate between the two is focused on different technical features that very quickly delve into minutia. However, the actual decision between the two should be made based on the type of traffic a site has.

Where's the traffic coming from?

If the majority of a site's traffic is side door traffic from Google, Facebook, and Twitter, the site should embrace mobile web and HTML5. Since most of the site's users are arriving via links, the content must quickly load in the
mobile browser. Such sites include music lyrics sites such as our site MetroLyrics and other types of information look up sites.

If a majority of a site's traffic is direct but intermittent traffic--meaning users come directly to the site, but only once in a while--the site should implement HTML5 mobile Web. These types of sites are "tourist sites" that are not visited regularly by people and therefore users are very unlikely to download an app. Such sites include corporate websites such as my company's homepage.

If the majority of a site's traffic is direct traffic where people are regularly going straight to the site's home page from a bookmark or typing in the URL, the site should use native apps. Such sites include, CNET Reviews, and other types of highly branded destination sites.

Sites with direct traffic that is intermittent--meaning people drop by every now and then--should still use HTML5 rather than native. For sites with a lot of direct traffic, native apps also provide useful additional features such as push notifications and offline storage, which are not relevant to sites with intermittent or side door traffic.

Sites that have an even mix of direct and side door traffic should also implement both native apps and an HTML 5 mobile view. A word of caution, however: there is always an inclination to heavily promote your native app to everyone going to your mobile Web site by forcing users to click through a native app promotion. This is a way to piss people off. Most of those visitors are clicking on a link in Google or Facebook and expect to see the content. They don't want to download your app.

What can you spend?

Once you determine whether to build an HTML5 mobile Web site or a native app, the next big question is how much you are willing to spend. Really, there are only two choices: complete and cheap or custom and expensive.

Sites should generally start with a turnkey and cheap solution. For turnkey mobile Web HTML5, vendors like Pressly and Mobify will take your content and make it sport a sexy, Flipboard-stye tablet interface. Wordpress includes mobile plugins that work great on
iPhone and
Android. Be sure to add a "view full site" option so that your users can opt out of the mobile experience and access functionality that the turnkey HTML5 solutions do not yet support.

To deliver turnkey native apps, services like MobileRoadie will consume your content, social feeds, and more and let you style good iPhone and Android native apps, with iPad soon to come. The apps are gorgeous and responsive and provide extensive options.

For the sites that need to support both mobile web and native apps, it is likely that the turnkey vendors will soon begin to support both distribution channels, and one vendor will be able to deliver best-of-breed solutions for both mobile HTML5 and native apps. For now, however, I suggest using a different vendor for each.

Once you have a baseline mobile presence, you can consider adding a custom experience that will support numerous features and user interface enhancements. Unfortunately, custom means expensive, both for HTML5 and native apps.

There are numerous systems integrators such as that deliver elegant, iPhone, iPad and Android native apps. Be aware that these integrators are going to need to be able to integrate with your registration, user profile, and content systems and that will likely require engineering and IT work. Some integrators such as FreeRange360 have an underlying platform that makes this type of customization relatively straightforward.

While HTML5 has come a long way, it is still not up to par with the native app experience. Some publishers, such as the Financial Times and Playboy, have come close to native app functionality by investing heavily in HTML5 in order to bypass Apple's 30 percent app store subscription fee. However, there are no turnkey JavaScript libraries that provide functionality such as efficient swiping and offline reading.
That said, it is relatively straightforward to efficiently deliver an excellent mobile Web experience. Libraries like jQuery mobile and Sencha mobile provide excellent HTML5 iPhone-style user interface controls, and it is easy enough in modern web frameworks such as PHP and Ruby to detect what type of device is requesting content and delivering a customized page for particular screen sizes, known as the "if viewport then" technique. It is tedious and cumbersome work, but can be done, and provides an excellent level of control and flexibility.

For properties that contain primarily text and images, you could consider a hybrid HTML5-native approach, where a mobile-optimized HTML5 site is wrapped with a native wrapper like PhoneGap. While this sounds like an ideal solution, consider that this approach is quite nascent, and that it takes quite a bit of work to make HTML5 work and look like a native app.

In summary, when discussing your mobile strategy, use the type of traffic your site has to determine whether to use HTML5 mobile Web or native apps, and then use your level of budget to decide whether to go turnkey or custom. And have some fun with your apps and please let me now what's worked for you.

Friday, February 10, 2012

How Sites Like MegaUpload Make Millions from Pirated Video

This post was also published on CNET and VentureBeat.

For the scope of this article, I am leaving all of the commentary on SOPA, PIPA, ACTA, and such aside for others much more well-versed than I to discuss.

A lot of people have been asking me the same question lately: Just how do sites like MegaUpload--recently taken down by an international collection of law enforcement--make hundreds of millions of dollars a year and fund lottery-winner style lifestyles that include mansions and private jets?

It's actually pretty straightforward. These sites use the same techniques as legitimate Web sites: search, social media, ad networks, and online payment processors.

Sites that feature links to illegal videos optimize for the keyword "links," and users that seek such videos have learned to search for "links." Generally, providers of legal content are not trying to land in searches for links; they are trying to land in searches for the word "videos." So searching Google for a popular TV show such as NCIS followed by the words "free links" returns sites that feature links to pirated copies of the TV show.

Clicking on one of these Google links, such as, returns a well-designed Web page featuring advertising from the likes of American Express and Hertz via retargeting ad networks such as Chango.

The page also features links to the latest NCIS episodes that are illegally hosted on download sites. Each link has a rating so that visitors will know when a file is no longer available due to a DCMA takedown. As there are numerous links available on numerous download sites, there is generally always an illegal copy available for viewing. It's technically legal to link to illegally hosted copyrighted content, so these types of sites are seemingly doing nothing wrong.

Like many profit-oriented Web sites, clicking on one of these links actually takes you to a secondary page so that the site can generate an additional page view.

Clicking on a link sends the user to file-sharing sites such as that allow streaming video of hosted files, including retargeted ad networks ads from advertisers such as Virgin America.

After watching a set amount of video, users are incentivized to pay to watch more using their credit cards, processed by payment networks such as Skrill. You're also offered the option to earn points towards viewing videos by accepting "offers," such as a Netflix trial subscription.

If this all looks familiar and resembles so many media sites you come across, they've succeded. That's the intention.

Thursday, February 02, 2012

Forget $100B! Facebook could soon be worth $200B

This post was also published on CNET and VentureBeat.

For all the naysaying about Facebook, that it's a flash in the pan and such, there are very few that say that "social" is going away.

Facebook has defined the social era of computing--and the companies that defined the previous eras of computing each command market values of $200 billion or more.

Facebook should get there, too.

IBM kicked off the mainframe era of computing and to this day is the leader in big enterprise computers and services. Microsoft was an early leader in personal computer software and now dominates microprocessor based desktops and servers. And after joining the scrum at the tail end of the dotcom boom of the 1990s, Google emerged as the leader of the Internet era of computing, amassing huge market share and most of Internet advertising's profits.

Coincidentally, those three companies--each of which dominated an era of computing--are now each worth roughly $200 billion.

While many still think of Apple as a computer company, it's not. It's reinvented itself as the leader of the mobile era of computing. Three quarters of Apple's revenues are now from the
iPhone, iPad, and
iPod, and it is in the process of re-creating the
Mac as a mobile computer with the MacBook Air. Apple, as the leader of the mobile era of computing, is now valued at an astounding $400 billion plus.

Given the history of IBM, Microsoft, Google, and Apple, that each respectively led the mainframe/minicomputer, personal computer, Internet, and mobile eras of computing, it is not that much of a stretch that Facebook, as the company that defines the social era of computing, will be right up there with them. Each of these eras have produced prodigious revenue and earnings, and as Facebook's S1 filing shows, social is already well on its way to stellar revenue and earnings--making the bulk of its money the way Google does, through advertising.

The leaders of each era have managed to lock in a generation of users ranging from business datacenters, PC operating systems and applications, and the portal by which people search. Facebook's social graph will be just as persistent. While other niche social networks will emerge, most of us aren't going to switch.

Of course, Facebook's current revenue and earnings does not justify such a valuation, but Facebook is still young and doubtless will figure out plenty of ways to make more money, including selling valuable new ad units such as sponsored stories against its increasing number of mobile users. So long as it it continues on its existing trajectory, the leader of the social era of computing will join others in the $200 billion club.

Or--dare I suggest?--the $400 billion club.